What are the uses of MEG?

Being the only brain imaging technic combining high anatomical and time resolutions, MEG has numerous applications in clinical and neuroscience research: Early diagnosis of Alzheimer’s disease, rehabilitation following strokes, diagnosis of mental disorders, and autism. This technique is also used to study brain dynamics, brain cognitive processes and consciousness.

MEG can be clinically helpful to localise the brain areas where a seizure originates and where epileptic activity spreads (epileptogenic brain area). 1/3 of the 50 million patients suffering from epilepsy are unable to be treated with medication, these patients are candidates for surgery, and with the use of MEG providing the the neurosurgeon with the localisation of the brain region to be resected, this technique can be shown to significantly improve the outcome of epilepsy surgery.

MEG is also used clinically to map functional areas of the brain before a tumour resection, or before an epilepsy surgery in order to delineate the functional brain regions that have to be preserved during the surgery and optimise the resection for achieving the best outcome. Brain tumours affect 1.6 million people per year.

Another emerging clinical application is mild Traumatic Brain Injury (mTBI) which affects over 60 million people a year, of which 20% will go on to develop permanent disabilities. 90% of diagnosis is currently completed based on the evaluation of external symptoms alone. X-Ray scanners and MRIs are not sensitive enough to detect mTBI. MEG has been shown to be able to detect tiny changes in brain function and connectivity, not visible in MRI, and thus provides an objective and specific measurement of the brain function and its changes, making it possible to do an objective diagnosis of mTBI. MEG specificity has been evaluated to 95% for mTBI. MEG is already used in clinics worldwide and raises a strong interest in the sports and military industry, with people particularly affected by mTBI. It’s a really relevant and promising clinical application of MEG.

MEG is also useful in neurofeedback and rehabilitation approaches, in drug development by quantifying drug efficiency and localising this effect in the brain. MEG can be used in the early phases of drug development to screen the pharmacological molecules that are good drug candidates. 

Other applications of MEG are expected to emerge in the upcoming years as neuroscience and clinical research progress.

Cryogenic MEG and its Limitations?

MEG is historically based on cryogenic sensors named SQUID's that are very sensitive but require cooling to ~4K (–269°C). In turn, this requires bathing these sensors in liquid helium and maintaining a vacuum between the sensors and the participant's scalp for thermal insulation.

Cryogeny imposes strong constraints and leads to high purchase and operating costs. Sensors have to be enclosed in a rigid helmet with thermal insulation and arranged in a fixed array around the scalp. These design concerns underpin the majority of MEG's limitations;

1. Because of the fixed array, participants must remain stationary in relation to the sensors throughout data collection. Consequently, coping with the MEG setting is difficult for some individuals;
2. The requirement for thermal insulation between the scalp and the sensor restricts proximity (the closest a sensor can get to the scalp is 2 cm), hence reducing signal strength;
3. The MEG signal strength is consequently reduces as it diminishes with the square of the distance from the source (the inverse square law);
4. The demand for stiffness also necessitates the adoption of a MEG helmet designed to fit 95% of people. In practice, this implies that the helmet is built for an individual with a larger head; the helmet will not fit properly on the majority of individuals, and the distance between the scalp and the helmet will vary across the head. This leads to disparate coverage. This impact is magnified for people with smaller heads, making it difficult to achieve both uniform covering and great sensitivity in infants;
5. Scanners are expensive to purchase and maintain because of the system's complexity, and the need for cryogenics requiring either a constant supply of liquid helium or a local helium re-liquefier.

The advantages of ⁴He OPM MEG?

Overall, Optically Pumped Magnetometer (OPM) sensors have various advantages over traditional SQUID MEG scanners, including:

• Increased signal sensitivity
• Improved spatial resolution
• Greater uniformity of coverage
• Lifespan compliance
• Freedom of participant mobility during scanning
• Lower operating costs and no need for supercooling
• Reduce system complexity

Helium Optically Pumped Magnetometers (⁴He OPMs) are the latest advanced OPM sensors that work at room temperature, i.e. they don’t require to be heated (as their counterpart alkali OPM) nor cooling as SQUIDs.

⁴He OPMs don’t require any cryogenic fluid, drastically reducing the operating cost of MEG and allowing for more affordable MEG devices for researchers and hospitals. They can be worn over the scalp, without any heating or burning discomforts and the patient is able to move during the recording thanks to the large He OPMs dynamics. ⁴He OPM MEG is particularly well suited to babies and children (small head sizes) as the sensors are worn over the scalp. The minimised distance scalp sensor also optimises the brain activity recording. Thanks to their large bandwidth,

⁴He OPM also allows for recording all the physiological and pathological brain activities, even with very high frequencies.

The technological breakthrough brought by ⁴He OPM fully reshapes the use of MEG and can help clinicians and researchers to facilitate the emergence of current clinical applications of MEG such as epilepsy, pre-surgery mapping or mTBI but also new applications such as early diagnosis of Alzheimer’s disease.

Optically Pumped Magnetometer (OPM) sensors have various advantages over traditional SQUID MEG scanners, including:

• Increased signal sensitivity
• Improved spatial resolution
• Greater uniformity of coverage
• Lifespan compliance
• freedom of participant mobility during scanning
• Lower operating costs and no need for supercooling
• Reduce system complexity

In spite of delivering high-quality spatiotemporal maps of electrophysiological activity, the existing MEG apparatus is constrained by complex field sensing technologies, resulting in significant limitations to application.

MEG is based on the measurement of magnetic fields generated (mainly) by synchronous dendritic current flow through neuronal assemblies outside the head. Mathematical modelling of these fields permits the development of three-dimensional images (termed source localisation) depicting the moment-to-moment variations in electrical activity as the brain responds to different experimental circumstances or cognitive demands.

The current generation of MEG scanners is hampered by substantial limitations that diminish their utility.

Current MEG systems (which are housed in a magnetically shielded environment to suppress background fields) employ pick-up coils that are coupled to superconducting quantum interference devices (SQUIDs) in order to gain sufficient sensitivity to measure the small (100 fT) magnetic fields generated by the brain. Typically, these sensors require cooling to ~4K (–269°C). In turn, this necessitates bathing the sensors in liquid helium and maintaining a vacuum between the sensors and the participant's scalp for thermal insulation. This requires sensors to be arranged in a fixed array around the scalp. These design concerns underpin the majority of MEG's limitations;

1. Because of the fixed array, participants must remain stationary in relation to the sensors throughout data collection. Consequently, coping with the MEG setting is difficult for some individuals;
2. The MEG signal strength diminishes with the square of the distance from the source (the inverse square law);
3. The requirement for thermal insulation between the scalp and the sensor restricts proximity (the closest a sensor can get to the scalp is 2 cm), hence reducing signal strength;
4. The demand for stiffness also necessitates the adoption of a MEG helmet designed to fit 95% of people. In practice, this implies that the helmet is built for an individual with a larger head; the helmet will not fit properly on the majority of individuals, and the distance between the scalp and the helmet will vary across the head. This leads to disparate coverage. This impact is magnified for people with smaller heads, making it difficult to achieve both uniform covering and great sensitivity in infants;
5. scanners are expensive to purchase and maintain because of the system's complexity, and the need for cryogenics requiring either a constant supply of liquid helium or a local helium re-liquefier.

In recent years, a novel magnetic field sensing method has emerged in the MEG field. OPMs are magnetic-field sensors with equivalent sensitivity to SQUIDs that do not require cryogenic cooling. This has resulted in the development of novel MEG systems, and 'OPM-MEG' scanners are beginning to outperform the existing state of the art in terms of data quality, uniformity of coverage, motion robustness, and system complexity, while being a relatively new technology.

Many of these obstacles are starting to be removed by the next MEG technology generation. "OPM-MEG" has the potential to drastically surpass the existing state of the art by utilising quantum sensors that are known as optically pumped magnetometers (OPMs). This could result in improved data quality (higher sensitivity and spatial resolution), adaptation to any head size or shape (from babies to adults), motion resilience (participants can move freely during scanning), and a less sophisticated imaging platform (without reliance on cryogenics).

Applications of OPM-MEG

At the core of the ⁴He OPM, there is a cell containing Helium atoms that possess spin and produce a magnetic moment, according to the laws of physics. The magnetic moments of these atoms are arbitrarily aligned in the absence of any outside influence. Moreover, He atoms, in their current state, are not sensitive to a magnetic field. The magnetic properties of He atoms are obtained in their metastable state, by applying a short discharge. There is no heating and therefore, no heat dissipation with ⁴He OPM.

A linearly laser pumping with a wavelength resonant to the D₀ quantum state transition (1083 nm for He atoms) brings macro properties, aligning all the magnetic moments of He atoms. Once in this state (spin alignment), He atoms do not absorb photons.

If a magnetic field (such as the field from the brain), impinges on the cell, it induces slight spin rotation, and therefore light absorption.

A photodetector is used to collect the photons and quantify the light absorption, which is proportional to the brain's magnetic field.

A first LF field at frequency ω translates the DC absorption properties to ω, and provides a directional measurement of the magnetic field. A second LF field at frequency Ω translates the DC absorption properties to Ω, and provides a second measurement axis of the magnetic field. The third measurement axis of the magnetic field is obtained by mixing ω  +/- Ω of the two first ones.

More detailed information about 4He OPM and they are working can be found in (Fourcault et al., Helium-4 magnetometers for room-temperature biomedical imaging: Toward collective operation and Photon-noise limited sensitivity 2021) and in (Beato et al., Theory of a He4 parametric-resonance magnetometer based on atomic alignment 2018).

Difference Between Alkali OPM and ⁴He OPM

	Alkali OPM	4He OPM
Power dissipated by a 50 sensors net	35 W	1W
Sensors dimensions	16.6 x 20 x 27 mm	19 x 19 x 50 mm
Bandwidth	1 – 130 Hz	DC – 2000 Hz
Dynamic range	5 nT	Up to 200 nT
Sensor Noise	15 fT/√Hz on 2/3 axis	<25 fT/√Hz on 2/3 axis
Accuracy	Closed loop on 3 axis	Closed loop on 3 axis
Reliability	Instability due to embedded-heater, laser source and alkali atoms reactivity	Very high - Only gas cell and optical length into sensors.
Operating conditions	Heating process (150°C)	Room temperature
Sensors localisation	Rigid Helmet: Self-localised from factory	Flexible Helmet: auto localised
Lifetime	2 years	~10 years

Literature on OPM-MEG

OPM-MEG can be used to measure a number of electrophysiological events that are often reported by MEG and EEG. Commonly, evoked responses to sensory stimuli of many modalities are evaluated, with OPMs giving high-fidelity measurements (Labyt et al., Magnetoencephalography with optically pumped 4he magnetometers at ambient temperature 2019), (Borna et al., Non-invasive functional-brain-imaging with an OPM-based magnetoencephalography system 2020).

Likewise, brain oscillations have been detected over many frequency bands with good SNR (Iivanainen et al., Potential of on‐scalp meg: Robust detection of Human Visual gamma‐band responses 2019). 

Using wearable OPM-MEG sensors, epileptiform activity has been characterised, suggesting the potential for future clinical application (Feys et al., On-scalp optically pumped magnetometers versus cryogenic magnetoencephalography for diagnostic evaluation of epilepsy in school-aged children 2022), (Feys et al., Recording of ictal epileptic activity using on‐scalp magnetoencephalography 2022).

Even with a lower channel count, whole-head system studies have demonstrated that wearable MEG performance can surpass that of conventional systems. Recent research compared a 50-channel wearable OPM-MEG system with favourable findings to a 275-channel cryogenic MEG device. (Hill et al., Multi-channel whole-head OPM-Meg: Helmet Design and a comparison with a conventional system 2020). Another study also quantified how a whole-head OPM MEG system could drastically improve performances in terms of brain activity recording compared to a cryogenic MEG system (Zahran et al., Performance analysis of optically pumped 4HE magnetometers vs. conventional SQUIDs: From adult to infant head models 2022).

With the emergence of wearable, whole-head devices, OPM-MEG is now possible to evaluate functional connectivity (Boto et al., Measuring functional connectivity with wearable MEG 2021), with electrophysiological networks being readily identifiable. The use of wearable MEG for brain-computer interface (BCI) has also been demonstrated (Wittevrongel et al., Practical real-time Meg-based neural interfacing with optically pumped magnetometers 2021): in a 'mind-spelling' exercise, participants were instructed to look at a virtual keyboard and fixate their gaze on the letter they desired to type. The OPM-MEG signals were then analysed by a machine learning algorithm to determine which letter the participant was observing. In 97.7% of trials, letters were properly identified. This not only highlights the possibility for future application in BCI, but also demonstrates the high-fidelity data that can be collected using OPM-MEG devices.

The performance of OPMs at low frequency is a limiting factor; there are three reasons for this. Initially, the intrinsic OPM sensor noise increases with decreasing frequency. Second, OPMs are magnetometers, which are sensitive to distant interference sources, and the interference is more difficult to shield the lower the frequency. Thirdly, for wearable systems, movement, even in extremely low background fields, will generate certain artefacts, which typically emerge at low frequencies. In recent research, however, a wearable OPM system was successfully used for cortical tracking of speech and the MEG signals tracked the rhythmicity of phrases (0.2–1.5 Hz signals) and words (2–8 Hz signals) with reconstruction accuracy comparable to that of previously reported in conventional MEG studies, suggesting that OPM-MEG is well suited to measure brain activity at frequencies below 4 Hz. Similarly, theta (4–8 Hz) oscillations in the hippocampus were effectively quantified utilising a novel design of OPM-array in which an OPM was inserted in the participant's mouth to boost array sensitivity, demonstrating the versatility of array design as well as sensitivity to low frequencies. (Tierney et al., Mouth Magnetoencephalography: A unique perspective on the human hippocampus 2021).

OPM-MEG systems have demonstrated distinct advantages over traditional MEG equipment. The benefits can be roughly divided into four categories:

1. Data quality: OPMs detect a MEG signal that is greater in amplitude and better spatially localised than SQUIDs due to the increased proximity of sensors to the scalp surface.
2. Adaptability: OPM-MEG can be customised to the head size and shape of individual participants, and the sensor array can be adjusted flexibly based on the requirements of particular investigations. In contrast to traditional MEG, OPM-MEG has the ability to adapt to individuals of any age, making these benefits especially significant in paediatric imaging.
3. Motion robustness: the capacity to scan people while they are in motion will allow data acquisition in participants who cannot handle the rigours of existing functional imaging environments, as well as the application of innovative experimental designs that are not achievable with traditional MEG or MRI.
4. System simplicity: the lack of dependency on cryogenic sensing enables simpler, cryogen-free instrumentation.

Additionally, OPM-MEG has advantages over other functional neuroimaging techniques: fMRI, for instance, is limited to haemodynamic measurement, has a poor temporal resolution, requires participants to be in a confined and loud area, and requires participants to remain still, making it difficult to quantify brain activity in naturalistic circumstances. While EEG and fNIRS provide realistic activity while scanning, they have either limited spatial (EEG) or temporal (fNIRS) resolution (fNIRS). For these reasons, within the landscape of functional imaging, OPM-MEG is beginning to stand out as an emergent tool that surpasses current technologies in a number of respects.

There are numerous sectors that stand to benefit from these advantages.

For instance:

• Greater spatial accuracy and sensitivity will be extremely beneficial for all functional mapping studies, including clinical applications (e.g., mapping epileptiform activity) and fundamental research. The scanning of neonates, infants, and children is easier with OPM-MEG than with SQUID MEG, which offers both clinical and fundamental research prospects (e.g., the study of neurodevelopment problems) (e.g., examining how electrophysiological activity and connectivity change during the early years of life).
• The ability to scan while free movement makes MEG accessible to individuals who would find it difficult to endure a traditional scanner, such as those with motor, tremor or behavioural problems. Additionally, motion tolerance gives the opportunity for novel forms of exploration (e.g., immersive environments, or naturalistic scenarios).
• OPMs are not restricted to the study of the brain; they are also being used to evaluate the electrophysiology of the peripheral nervous system, muscles, heart, and even enteric nervous system.

OPM-MEG is a developing technology that has the potential to offer neuroscientists a noninvasive, one-of-a-kind window into brain function. Even now, OPM-MEG has considerable advantages over previous techniques, and while technical obstacles remain, there do not appear to be any fundamental obstacles to the technology's ongoing growth and improvement. This will certainly provide new and intriguing opportunities for future neuroimaging-based research of brain function as the discipline evolves.

Magnetoencephalography with Helium optically pumped magnetometers (OPM-MEG): the next generation of functional neuroimaging.